Introduction

VILD is (S)-1-[N-(3-Hydroxy-1-adamantyl)glycyl] pyrrolidine-2-carbonitrile (Fig. 1). VILD is a relatively new anti-diabetic drug used for controlling patients with type 2 diabetes mellitus. VILD subsequently works by inhibiting glucagon-like peptide-1 (GLP-1) and gastric polypeptide inhibitor (GIP) inactivation by DPP-4 [1, 2].

Fig. 1
figure 1

Chemical structure of VILD drug

Different analytical methods have been described for the quantification of VILD from biological samples, including ultraperformance liquid chromatography–mass spectrometry (UPLC–MS/MS) [3], high–performance liquid chromatography (HPLC) [4], atomic absorption and emission spectrometry [5], capillary electrophoresis (CE–MS/MS) [6], LC–MS/MS [7, 8], and hydrophilic interaction chromatography (HILIC/MS/MS) [9] methods, as well as in commercial formulations, including gas chromatography (GC–MS) [10], ultraviolet (UV) spectrophotometry [11,12,13,14], HPLC [15,16,17], reversed-phase (RP–HPLC) [18,19,20], and UV–HPLC [21] methods. Even though these chromatographic methods show high sensitivity and accuracy, they have several disadvantages such as the need for more laborious pretreatment of the sample in some cases, a long analysis time, and the use of highly toxic reagents or organic solvents. Besides, most of them are indirectly based on the formation of stable ion–pair complexes/colored chromogens prior to the analysis. Electroanalytical methods can present some advantages in comparison with chromatographic methods, such as lower reagent consumption, shorter analysis time, lower instrumentation cost, and the use of low-toxicity reagents (usually aqueous buffer solutions).

To the best of our knowledge, there have been two reported methods for determining VILD in commercial formulations and serums by voltammetric methods [22, 23]. The first method is based on the determination of VILD using a pencil graphite electrode (PGE) in tablets and human urine using square wave voltammetry (SWV) with a limit of detection (LOD) equal to 82 nmol L−1 [22]. The second method is based on the determination of VILD in biological fluid and drug forms using platinum (Pt) and glassy carbon (GC) electrodes using cyclic voltammetry (CV) and linear scanning voltammetry (LSV) techniques with a LOD equal to 0.24 mmol L−1 [23]. PGE has advantages such as being environmentally friendly, inexpensive, disposable, not requiring cleaning during measurement, and having very good analytical performance [22]. Aside from the majority of them, the performance of PGE in terms of LOD is less sensitive than the values obtained with the modified electrodes in the proposed method (the present work). In addition, the cost of the GC electrode and the necessity to clean the electrode surface before each measurement can be considered as disadvantages in comparison with the modified electrodes in the present work. When the analytical data obtained in this study are evaluated, it is seen that the performance of modified electrodes is much more sensitive in terms of LOD than the values obtained with PGE [22] and GC/Pt [23].

On the other side, carbon-paste electrodes (CPEs) (CP is a mixture of an electrically conducting graphite powder and a pasting liquid) have been widely used as a suitable matrix for the preparation of modified electrodes due to their low cost, simple preparation, and compatibility with various types of modifiers but low background current. However, most of the recent studies on CPE [24, 25] revealed that its sensitivity is relatively poor. This may be attributed to its high surface hydrophobicity, which is mainly responsible for the high overpotential (irreversibility) and sluggishness of the kinetics of the electrode process, resulting in weak electrochemical responses [26, 27].

Recently, to improve the sensitivity, selectivity, detection limit, and other features of CPE, chemically modified carbon paste electrodes (CM/CPEs) such as Ca-MMT have been used due to their significant advantages, such as increasing peak current and decreasing overpotential for redox systems.

Ca-MMT can play an important role as the parent host structure for inorganic and organic compounds due to their intrinsic properties such as high surface area, high electro-catalytic activity, thermal stability, and ion exchange [28,29,30], and the presence of hydroxyl group active sites on their interlayer surfaces [31, 32]. As a result, MMT has been widely used for modifying electrodes in order to improve their determination sensitivity toward anions [33], cations [34,35,36], and organic compounds [35, 37, 38]. Moreover, it has already been demonstrated that this modifier improves the electron transfer kinetics and the magnitude of the analytical signal when used as an electrochemical sensor [36, 37]. Consequently, MMT clay is a promising electrode for the determination of the VILD. Gosh and Bard were the first to describe clay as an adsorbent that may be combined with carbon [39]. There are no previous reports on the usage of a CPE modified with Ca-MMT clay for the determination of VILD in pharmaceutical formulations and biological samples. The mean limit of detection of VILD using a CPE modified with Ca-MMT clay was calculated to be 0.285 nmol L−1 in bulk form compared to the reported electrodes [22, 23]. The proposed method has been successfully applied to pharmaceutical formulations and biological samples.

Experimental

Apparatus of electrochemical measurements

A computer-controlled potentiostat model 273A-PAR (Princeton Applied Research, Oak Ridge, TN, USA) with the software 270/250-PAR was used for the voltammetric measurements. A micro-electrolysis cell containing three electrodes, namely Ca-MMT clay modified with CPE as the working electrode, Ag/AgCl as the reference electrode, and platinum wire as the counter electrode, was used for the voltammetric measurements. A magnetic stirrer was used to provide the convective transport during the preconcentration step.

Apparatus of spectroscopy measurements

Fourier-transform infrared (FTIR) spectra were reported on Bruker model Tensor 27 FTIR spectrometer using KBr pellets. FTIR spectra at room temperature were recorded in the range of 4000–400 cm−1, and the band intensities were expressed in terms of transmittance (%).

The scanning electron microscope (Quanta) instrument model (FEG-250) was used to show the difference in the morphology of carbon paste (CP), Ca-MMT, and the modified CP electrodes with various ratios of Ca-MMT clay at the voltage of 10 kV.

Reagents and solutions

Materials

Galvus@50 mg contains VILD with excipients as anhydrous lactose per tablet, which was purchased from Global Napi Pharmaceuticals (Cairo, Egypt). Phosphoric acid (H3PO4), boric acid (H3BO3), sodium hydroxide (NaOH), potassium nitrate (KNO3), methanol (CH3OH), and glacial acetic acid (CH3COOH) were gifted from (Sigma-Aldrich). Potassium nitrate (KNO3) and methanol (CH3OH) were gifted from (Sigma-Aldrich). Distilled water (AR Grade, S. D. Fine Chemicals Ltd., Mumbai, India) and Whatman filter paper no. 1005–090 (pores size 2.5 micron and diameter 90 mm) were used in the study.

Standard bulk VILD solution

A stock standard solution of 1.0 mmol L−1 bulk VILD was freshly prepared in the binary mixed solvents (methanol: water) (50:50, v:v) used as a solvent. The required concentrations (1 × 10–4 to 1 × 10−9 mol L−1) of VILD were prepared from the stock standard solution. A previous publication reported that VILD was partially hydrolyzed in an aqueous solution to a carboxylic acid metabolic product, M 20.7 (LAY151), after being stored at a certain temperature for several weeks. For this reason, all solutions were kept in the dark in a refrigerator and were used within several hours to avoid hydrolysis. All experiments were carried out at the ambient temperature of the laboratory (25 ± 2 ℃).

Supporting electrolyte

A series of BR universal buffers (0.04 M acetic acid, boric acid, and phosphoric acid, pH 4.0–10) were used as supporting electrolytes. The pH of the buffer solutions was adjusted with 4 mol L−1 NaOH and 4 mol L−1 HCl solutions. A pH-meter (model AD 111, Romania) of accuracy ± 0.02 and resolution 0.01 pH units was used for the pH measurements. Deionized water was supplied from a Purite-Still Plus de-ionizer connected to an AquaMatic double-distillation water system (Hamilton Laboratory Glass LTD, Kent, UK).

Solutions of pharmaceutical formulations and procedure

Galvus® tablets (obtained from Global Napi pharmaceuticals (Cairo, Egypt)) labeled to contain 50 mg VILD per tablet were quantitatively weighed, and the average mass per tablet was determined. Six tablets were then grounded in an agate mortar to a homogeneous fine powder. An adequate of this powder equivalent to the weight of one tablet was accurately transferred into a 100 mL volume calibrated flask containing 50 mL of binary mixed solvents. The content of the flask was sonicated for about 5 min and then filled up with binary mixed solvents (methanol: water) (50:50, v:v). The solution was then filtered through the Whatman filter paper. The sample from the clear supernatant liquor was withdrawn. A known volume of the sample was added into a voltammetric cell, and SW voltammograms were recorded. The nominal content of VILD in each tablet was determined by referring to the related regression equations. The calibration curve for SWV analysis was constructed by plotting the peak current against the potential.

Solution of spiked human serum and procedure

Serum was obtained from healthy donors and used shortly after collection. In each of eight centrifugation tubes containing a certain concentration of the bulk VILD under investigation, 1 mL of the human serum sample was transferred, and the volume was then made up to 2 mL with methanol (as a precipitating agent of proteins). The precipitated proteins were separated by centrifugation for 5 min at 14,000 rpm. The clear supernatant layer of each solution was filtered through a Whatman filter paper to produce a protein-free spiked human serum with different concentrations of the bulk VILD (10−8–10−4 mol L−1). An aliquot of this solution (0.01–0.1 mL) was diluted to 10 mL with a BR universal buffer and then transferred into a dark voltammetric cell.

Working and modified electrode fabrication

Carbon paste electrode (CPE)

The CP was prepared by thoroughly hand mixing 3.0 g of graphite powder (1–2 microns) from Sigma-Aldrich with 1.08 mL of mineral oil from Merck in a mortar with a pestle. The CP was packed firmly into the hole of the electrode rode, and the electrode surface was smoothed by polishing using a piece of butter paper until it was shiny. The electrode body was constructed by a small Teflon rode (4.0 mm internal diameter and 1.5 mm deep), and a thin wire of a copper screw pushing the paste was inserted through the opposite end and acting as an electrical contact. An electrochemical pretreatment of this electrode was done by several sweep runs to obtain a low background current using a CV between 0.0 and 1.4 V in a BR universal buffer of pH 7.0.

Carbon paste electrode (CPE) modified with Ca-MMT clay

The appropriate ratio of Ca-MMT/graphite was calculated using SW-AdASV. A series of Ca-MMT from Sigma-Aldrich in different masses was added to the different masses of graphite, with the total weight of the composite being 2.0 g. To prepare a homogenous 5% (w/w) Ca-MMT-modified CP, an amount (1.90 g) of graphite powder and 0.10 g of Ca-MMT clay were mixed uniformly by milling in an agate mortar. Then, 0.72 mL mineral oil was added and milled again to give a homogenous 5% (w/w) Ca-MMT-modified CP. Variously modified carbon pastes containing different mass percentages of Ca-MMT clay (1, 3, 7, and 9%, w/w) were similarly prepared.

Result and discussion

Electrochemical and morphological characterization of the proposed working electrode

Cyclic voltammograms of 1.0 µmol cm–3 K3[Fe(CN)6] at different scan rates in 0.1 mmol cm–3 KCl were recorded on both types of electrodes, namely CPE and 5% Ca-MMT/CPE (Fig. 2A, B). For a reversible process, the Randles–Sevcik formula has been used [40]. The peak current (Ip) was linearly proportional to ν 1/2 as follows:

$$I_{{{\text{pa}}}} {\mkern 1mu} = {\mkern 1mu} 2.69{\mkern 1mu} \times {\mkern 1mu} 10^{{5}}n^{{3/2}} {\text{ A }}C^{{\text{*}}} {\text{ }}D^{{1/2}} \nu ^{{1/2}}$$
(1)

where Ipa is the anodic peak current, n is the number of electrons (n = 1), A is the electroactive surface area of the electrode (cm2), C* is the concentration of potassium ferricyanide (mol cm–3), D is the diffusion coefficient (7.6 µcm2 s−1), and ν is the scan rate (V s–1). From the slope values (0.3208 × 10−3 and 0.4906 × 10−3) of the plot of Ipa vs ν1/2, the area of the electrode surface was calculated to be 0.4310 ± 0.0001 and 0.6600 ± 0.0002 cm2 for CPE and 5% Ca-MMT/CPE, respectively (insets of Fig. 2A, B). The peak-to-peak separation (ΔEp) of cathodic and anodic signals of model redox system K4[Fe(CN)6]/K3[Fe(CN)6] at 80 mV s–1 was calculated to be 289 and 194.2 mV for CPE and 5% Ca-MMT/CPE, respectively (Fig. 2C). A lower peak separation and higher surface area of 5% Ca-MMT/CPE than CPE were observed, indicating that 5% Ca-MMT/CPE had higher conductivity, electro-catalytic activity, and electron transfer rate than CPE.

Fig. 2
figure 2

Cyclic voltammograms for 1.0 µmol cm–3 K3[Fe(CN)6] in 0.1 mmol cm–3 KCl at different scan rates on A CPE and B 5% Ca-MMT/CPE; C cyclic voltammograms of 1.0 µmol cm–3 K3[Fe(CN)6] recorded at (a) bare CPE, (b) 5% Ca-MMT/ CPE in 0.1 mmol cm–3 KCl at a scan rate of 80 mV s–1. The insets show the anodic and cathodic peak currents (Ip (a, c)) vs ν 1/2 plots for A CPE and B 5% Ca-MMT/CPE

On the other side, the response stability of the 5% Ca-MMT/CPE was investigated for a 1.0 µmol cm–3 K3[Fe(CN)6] solution in 0.1 mmol cm–3 KCl by performing numerous CV measurements at a fixed potential scan rate of 80 mV s−1. Thus, 40 cyclic voltammograms were recorded (Fig. S1). The cyclic voltammograms obtained on the 5% Ca-MMT/CPE remained practically constant during the various potential cycles. The voltammograms showed that the ΔEp values did not change. Also, the obtained relative standard deviation (RSD %) values were 1.47 and 1.12% for the anodic and cathodic peak currents, respectively. These results constitute a demonstration of the high stability of 5% Ca-MMT/CPE response.

As shown in Fig. 3, scanning electron microscopy (SEM) was used to characterize the bare CPE morphology and that modified one with Ca-MMT clay in different ratios. The morphological surface studies of Ca-MMT clay and modification of bare CP electrode using 5% and 7% (w/w) Ca-MMT clay were carried out by SEM. Graphite has a layered structure that consists of stacked sheets of carbon atoms could be seen in the SEM image of the bare CPE (Fig. 3A). In Fig. 3B, SEM for Ca-MMT clay has been seen as plate crystal and clusters of sheets, similar to smectite clay minerals.

Fig. 3
figure 3

SEM images of the surface of A CPE, B Ca-MMT clay, C 5% Ca-MMT/CPE, and D 7% Ca-MMT/CPE

SEM showed that 5% (w/w) Ca-MMT/CPE (Fig. 3C) had a discrete, spherical morphology with a smooth surface and small particles with internal holes (as illustrated by the arrows), meaning that it had a higher surface area than the bare CPE. These holes correspond to the contact between the agglomerates of particles. Consequently, the 5% (w/w) Ca-MMT/CPE was chosen for assay of the VILD throughout the work. SEM for 7% (w/w) Ca-MMT/CPE showed a network of densely packed amorphous particles with ill-defined shaped structure, as shown in Fig. 3D.

Structural characterization of the proposed working electrode

As shown in Fig. 4, the FTIR analyses of the CPE, Ca-MMT clay, and 5% Ca-MMT/CPE were taken in the range of 400–4000 cm−1. As shown in Fig. 4a, the spectrum of pure graphite had absorption peaks at 1639 and 1383 cm−1 that were related to the C=C stretching vibration of the benzene ring of graphite [41]. As shown in Fig. 4b, the spectrum of the 5% Ca-MMT/CPE was recorded and is similar to the spectrum of Ca-MMT clay.

Fig. 4
figure 4

FTIR spectra of a CPE, b 5% Ca-MMT/CPE, and c Ca-MMT clay

As shown in Fig. 4c, the MMT clay is characterized by a very strong multiple absorption bands [42], observed in the range of 3621 to 3449 cm−1. A single sharp band observed at 3621 cm−1 followed by a broad band at 3449 cm−1 in the mineral is assigned to OH stretching (υ1–OH) of structural hydroxyl groups and hydration of water present in the mineral [43, 44]. This indicates the possibility of the hydroxyl linkage between octahedral and tetrahedral layers. Also, the peak produced at around 3621 cm−1 probably shows the presence of magnesium in the structure of the analyzed Ca-MMT clay [45]. A very sharp and intense band observed at 1639.2 cm−1 corresponds to the asymmetric (υ2–OH) (deformation mode) of the interlayer water and a structural part of the mineral [46]. A very strong absorption band observed at 1045.23 cm−1 is assigned to the stretching vibration Si–O band, which is characteristic of layered MMT silicate mineral and assigned to the triple degenerate extension (υ3) Si–O (in the plane) in SiO4 tetrahedron. The absorption band at 1383.68 cm−1 is corresponded to the stretching vibration Si–O band (υ4) Si–O (out of the plane). The peak at 1108 cm−1 is the longitudinal Si–O stretching mode. An additional peak also observed at about 1456 cm−1, may be due to the presence of sodium carbonate, from which the absorption band at 2372 cm−1 may be due to the CO2− stretching vibrational of sodium carbonate [47]. The OH bending peaks (deformation mode) of Al2OH and AlMgOH point out the octahedral substitution process, which were observed at 915 and 797 cm−1, respectively [48]. The band observed at 721 cm−1 and a sharp band at 623 cm−1 indicated Si–O stretching due to the presence of quartz [49, 50]. The band observed at 522 cm−1 is corresponded to the deformation mode (bending vibration) of Si–O–Al group in octahedral, while the band at 465 cm−1 is assigned to the Si–O–Si deformation band (bending vibration) [51]. Based on these results, the Ca-MMT clay and 5% Ca-MMT/CPE have a strong adsorption property toward the VILD drug compared to graphite. This may be due to the more functional groups of the clay as the OH group that are able to react with charged species ions of the drug and give an enhancement in the electrical current.

Electrochemical behavior of the VILD

Study of the potential scan rate effect

As shown in Fig. 5A, the electrochemical behavior of VILD at the 5% Ca-MMT/CPE surface was investigated using CV in a BR universal buffer of pH 7.0 at various scan rates of 70–500 mV s−1. The cyclic voltammograms obtained for 1.0 µmol L−1 VILD exhibited well-defined irreversible anodic peaks onto the 5% Ca-MMT/CPE surface. These irreversible anodic peaks are attributed to the oxidation of the secondary amine group (NH) of the VILD at the 5% Ca-MMT/CPE surface. It was observed that the oxidation peak potentials of the VILD positively shifted with an increase in the scan rate in the range of 70–500 mV s−1 at pH 7.0 (Fig. 5A). As shown in Fig. 5B, the peak current increased.

Fig. 5
figure 5

A Cyclic voltammograms of 1.0 µmol L−1 VILD at various scan rates of 70–500 mV s−1 in a BR universal buffer of pH 7.0, B variation of Ip (μA) vs ν (mV s−1) of 1.0 µmol L−1 VILD, and C variation of log Ip (μA) vs log ν (mV s−1) of 1.0 µmol L−1 VILD onto the 5% Ca-MMT/CPE surface

linearly with the increasing scan rate over the range 70–500 mV s−1. A linear relationship between the Ip/μA vs v/mV s−1 is expressed by an Eq. (2):

$$I_{{\text{p}}} /\mu A{\mkern 1mu} = {\mkern 1mu} 0.0100\,\left( { \pm {\mkern 1mu} 0.0001} \right)v/{\text{m}}\,{\text{V s}}^{{ - 1}} {\mkern 1mu} + {\mkern 1mu} 0.31\,\left( { \pm {\mkern 1mu} 0.05} \right)\,\left( {R^{2} {\mkern 1mu} = {\mkern 1mu} 0.998} \right)$$
(2)

Based on the good linearity of the Ip/μA vs v/mV s−1 plot, suggested that the oxidation process was controlled by a combination of diffusion and adsorption processes [52,53,54,55,56].

Furthermore, a linear relationship was observed between the log (Ip/μA) vs the logarithm of the scan rates (v/mV s−1) (Fig. 5C), based on an Eq. (3):

$${\text{log}}\left( {I_{{\text{p}}} /\mu A} \right){\mkern 1mu} = {\mkern 1mu} 0.87\,\left( { \pm {\mkern 1mu} 0.01} \right)\,{\text{log}}\,\left( {v/{\text{m}}\,{\text{V s}}^{{ - 1}} } \right){\mkern 1mu} - {\mkern 1mu} 1.50\,\left( { \pm {\mkern 1mu} 0.03} \right)\,\left( {R^{2} {\mkern 1mu} = {\mkern 1mu} 0.997} \right)$$
(3)

The slope value of (log Ip vs log v) was 0.87. This slope was an intermediate value between the theoretical value of 0.5 for diffusion-controlled process and close to the theoretical value, of 1, for the adsorption process, confirming that the oxidation process of the VILD was controlled by a combination of diffusion and adsorption process [52,53,54,55,56].

For an adsorption-controlled process, the charge transfer coefficient (α) can be calculated from the variation in Epa with lnυ following the Laviron’s equation [57]; its corresponding regression equation was as follows:

$$E_{{{\text{pa}}}} {\mkern 1mu} = {\mkern 1mu} E^{{\text{o}}} {\mkern 1mu} + {\mkern 1mu} \left( {{\text{RT}}/\alpha nF} \right)\,{\text{ln}}\,\left( {{\text{RT}}k_{{{\text{et}}}} /\alpha nF} \right){\mkern 1mu} - {\mkern 1mu} \left( {{\text{RT}}/\alpha nF} \right)\,{\text{ln}}\upsilon$$
(4)

According to Nicholson and Greef [58,59,60], the values of αana (product of symmetry transfer coefficient, α, and the number of electrons, na, transferred in the rds) of (0.22–0.64) were estimated from the slope values of the obtained Ep vs lnυ plots as follow:

$$\left\{ {\left( {\Delta E_\text{p} \left( V \right)} \right)/ \Delta \ln \left( {\upsilon \left( {{\text{mV s}}^{{{ - }1}} } \right)} \right)\, = \, 0.0256/2\alpha_{a} n_{a} } \right\}$$
(5)

The most probable values of the anodic transfer coefficient αa (0.11–0.32) were estimated at various pH levels (4–10), at the number of electrons, na, transferred in the rds equal to 2.0. This confirmed the irreversible nature of the electrode reaction of the VILD at 5% Ca-MMT/CPE.

Study of the pH effect

Moreover, the oxidation peak potential of the main peak of VILD shifted to a less positive value with an increase in the pH levels from 4–10 (Fig. 6A), which indicated that the protons were involved in the electrode reaction process [61]. As shown in Fig. 6B, a linear relationship of the anodic peak potential (Epa) vs (pH) was obtained at a scan rate of 100 mV s−1 according to the following Eq. (6):

$$E_{{{\text{pa}}}} \left( {\text{V}} \right)\, = \, - \,0.06{\mkern 1mu} \left( { \pm 0.002} \right){\text{ pH}}\, + \,1.62{\mkern 1mu} \left( { \pm \,0.01} \right),{\mkern 1mu} \left( {R^{2} \, = \,0.991} \right)$$
(6)

A plot of Ep vs pH was found to give a slope value of 0.06 V pH−1 at a scan rate of 100 mV s−1. This value is very close to the theoretical value at 0.059 V pH−1, indicative of an equal number of protons and electrons involved in the oxidation step. The number of protons in the rate-determining step (rds) was investigated according to the Nernst equation for an irreversible process [62] based on the slope value of the obtained Ep vs pH of VILD at a scan rate of 100 mV s−1 as follows:

$$\left\{ {{\text{(}}E_{{\text{P}}} (V))/\Delta {\text{pH}}{\mkern 1mu} {\text{(0}}{\text{.061)}}{\mkern 1mu} } \right.{\text{ = }}\left\{ {\left( {0.059/\alpha } \right) \cdot ({\text{Z}}_{{{\text{H}} + }} /n_{a} )} \right\}$$
(7)

where na and ZH+ are the number of electrons and protons involved in the rds, respectively, and α is the symmetry transfer coefficient. It is well known that two electrons (na = 2) and one proton (ZH+  = 0.66 ≈ 1) are involved in the rds, i.e., the ratio (ZH+/ na) = 1/2. Accordingly, the α–value of 0.32 was estimated from the slope value (0.061 V pH−1) of the (Ep) vs (pH) plot, indicating the irreversible nature of the electrode process of the VILD at the 5% Ca-MMT/CPE surface.

Fig. 6
figure 6

A Cyclic voltammograms of 1.0 µmol L−1 VILD at a scan rate of 400 mV s−1 in a BR universal buffer of different pH levels and B variation of Ep– pH of 1.0 µmol L−1 VILD onto the 5% Ca-MMT/CPE surface at a scan rate of 100 mV s−1

Based on the results obtained in Fig. 6B, the presence of two linear zones (intersection plots) around pH = 9.0 was investigated. The break at pH = 9.0 is related to the pKa value of VILD; this is due to changes in the protonation–deprotonation equilibrium of the VILD drug. This pKa1 value is very close to the value of VILD in the literature that equal to 9.03 [21, 63, 64].

As shown in Scheme 1 , in acidic media (pH values < 7.0), the cationic form was expected (1) to form a basic VILD (2) (pKa1 of 9.0 is attributed to dissociation of protonated basic nitrogen).

While, in strong basic solutions, the hydroxyl group may be dissociated into anionic VILD (3) and the pKa2 value was expected to be 14.71 based on the literature [21, 64].

Scheme 1
scheme 1

Possible acid-base equilibrium of VILD drug

The possible reaction mechanism of the VILD drug onto the 5% Ca-MMT/CPE surface

Based on the results obtained for the electrons and protons number determination of VILD, the reaction mechanism was suggested onto the 5% Ca-MMT/CPE surface, as shown in Scheme 2. The electrochemical pathway of VILD was suggested based on the removal of the acidic alpha proton adjacent to the alpha carbon of carbonyl group (1) to form the carbanion (2).

This carbanion is followed by the removal of one electron to form the free radical (3). After that, the free electron of the carbon radical forms a π bond with nitrogen to form a nitrogen-free radical (4). This nitrogen-free radical is accompanied by the removal of one electron to form the quaternary ammonium Schiff base (5), which on losing a proton in the subsequent step forms the imine VILD (6) [22]. Generally, the mechanism was suggested based on the removal of two electrons and protons for VILD in the subsequent steps to form the imine VILD [65, 66].

Scheme 2
scheme 2

Suggested reaction mechanism for the oxidation of the VILD in the basic media onto the 5% Ca-MMT/CPE surface

The adsorption behavior of the VILD

As shown in Fig. 7, the adsorption character of the VILD drug onto the surface of 5% Ca-MMT/CPE in the BR universal buffer of pH 7.0 at various scan rates in the range of 100–500 mV s−1 was illustrated using the CV technique. The cyclic voltammograms for 100 nmol L−1 VILD in the BR universal buffer of pH 7.0 at different scan rates of 100 (Fig. 7A), 300 (Fig. 7B), and 500 mV s−1 (Fig. 7C) were recorded after an accumulation time (tacc) of 40 s.

Fig. 7
figure 7

Successive cyclic voltammograms of 100 nmol L−1 VILD in the BR universal buffer of pH 7 at different scan rates of A 100, B 300, and C 500 mV s−1 after an accumulation time of 40 s

In Fig. 7, the repetitive cyclic voltammograms show that the peak currents (curves a) decrease in the second (curves b) and third cycles (curves c) at different scan rates. These behaviors at different scan rates indicate an adsorption character of VILD onto the surface of 5% Ca-MMT/CPE.

In addition, as shown in Fig. 8, a plot of the peak current versus the scan rate within the range of 0.10–0.50 Vs−1 after an accumulation time of 40 s for VILD using an Eq. (8) [67, 68] gave a straight line relation with a slope of 5.10 × 10−6 A Vs−1.

Fig. 8
figure 8

Variation of Ip vs υ of 100 nmol L−1 VILD onto the 5% Ca-MMT/CPE surface in a BR universal buffer of pH 7.0 after an accumulation time of 40 s

Based on these results, Γ0 was found to be 3.0 × 10−12 mol cm−2 using an Eq. (8) [67, 68]:

$$I_{{\text{p}}} { }\, = \,{ }\left( {n^{{2}} F^{{2}} A\Gamma^{{0}} /{4}RT} \right)\upsilon$$
(8)

where Ip is the peak current in amperes, A is the surface area of the electrode (0.6600 ± 0.002 cm2), v is the scan rate in Vs−1, Γ 0 is the surface coverage of adsorbed reactant in mol cm−2, and n is the number of electrons participating in the reaction (n = 2).

On the other side, a sharp and symmetric anodic peak with a maximum peak current for 1.0 µmol L−1 VILD was investigated using a CV study at the modified CPE (5%Ca-MMT/CPE) (Fig. 9, curve b) compared to the peak current at an unmodified CPE (Fig. 9, curve a). Moreover, the oxidation of VILD onto the 5% Ca-MMT/CPE was carried out at less positive potential compared to the CPE, reflecting the high electrocatalytic efficiency of the Ca-MMT clay compared to the CPE toward the VILD.

Fig. 9
figure 9

Cyclic voltammograms of 1.0 µmol L−1 VILD recorded at (a) bare CPE and (b) 5% Ca-MMT/CPE in a BR universal buffer of pH 7.0 at a scan rate of 300 mV s−1 after an accumulation time of 40 s

The voltammogram of VILD recorded at the bare CPE exhibited a relatively poor oxidation peak compared to the 5%CaMMT/CPE under the experimental conditions, indicating that the electrochemical activity of VILD at the bare CPE is very poor and/or bare CPE possesses poor adsorption ability toward VILD (Fig. 9, curve a). Also, the enhancement of the peak current at 5%Ca-MMT/CPE (Fig. 9, curve b) compared to the CPE, this may be due to the high cation exchange capacity of Ca-MMT clay [28,29,30] or its high adsorption ability to VILD attributed to its well-layered structure, high surface area (0.660 cm2), smaller peak separation (194.2 mV), and the negative charge held by its platelets that promote electrostatic interactions with protonated VILD species. Since VILD is protonated and has a positive charge in a BR universal buffer of pH 7.0 according to the value of pKa1 in Scheme 1. It can be concluded that the synergistic effect of Ca-MMT clay can effectually increase the rate of VILD preconcentration from aqueous solution to the electrode surface by adsorption. Therefore, 5%Ca-MMT/CPE provides higher currents and thus higher electrocatalytic efficiency compared to the CPE for VILD. All these results were carried out in a BR universal buffer of pH 7.0 at a scan rate of 300 mV s−1 after an accumulation time of 40 s.

Optimization of the electroanalytical parameters

Effect of quantity of Ca-MMT clay

As shown in Fig. 10A and B, the different ratios (w/w) of Ca-MMT clay in the graphite paste remarkably influence the square wave anodic stripping voltammetry peak current magnitude of 100 nmol L−1 VILD compared to the bare CPE; at pH = 7.0, Eacc = 0.1, and tacc = 40 s. The stripping peak current magnitude firstly increased upon the increase of the Ca-MMT clay ratio up to 5% (w/w) in the CP and then decreased at higher clay ratios. Such an increase in the magnitude of the stripping peak current was expected due to the increase in the number of surface reactive sites. The presence of more clay resulted in more binding sites for the VILD on the MMT surface, leading to the improvement of the efficiency of preconcentration of the VILD at Ca-MMT. However, the conductivity of the modified CPE dropped with an increase in the ratio of the non-conducting Ca-MMT clay; hindering the electron transfer process and increasing the background current. These findings implied that the modifier of 5% (w/w) Ca-MMT/CPE was critical for the development of the VILD. The results are investigated at the frequency (f), scan increment (∆Es), and pulse amplitude (a) of 80 Hz, 10 mV, and 25 mV, respectively.

Fig. 10
figure 10

(A) SW-AdAS voltammograms of 100 nmol L−1 VILD in a BR universal buffer of pH 7.0, following preconcentration onto the bare CPE (a) and CPE modified with 1.0% (b), 3.0% (c), 5.0% (d), 7.0% (e) and 9.0% (f) (w/w) Ca-MMT clay by adsorptive accumulation for 40 s at 0.1 V and (B) its corresponding plot of Ip as a function of % (w/w) Ca-MMT clay; (f = 80 Hz, ΔEs = 10 mV and a = 25 mV)

Effect of the pH of supporting electrolyte

The pH of the electrolyte had a significant effect on the VILD drug. The influence of pH on the response of the 5% Ca-MMT/CPE was studied at 100 nmol L−1 VILD. The pH dependence of the VILD oxidation was investigated using the SW-AdASV approach. The BR buffer solution was used in the pH range of 4–10. The Ip is associated with the oxidation of the VILD and increased with the increase in the pH value from 4.0 to 7.0. At pH values higher than 7.0 the Ip decreased abruptly. The oxidation potential of the VILD was shifted to less positive values with an increase in the pH value, indicating that protons were involved in the oxidation mechanism. These findings implied that the neutral medium (pH 7.0) was critical for the development of the VILD.

Effect of pulse parameters

The resulting oxidation peak current of VILD using SW-AdASV at the 5% Ca-MMT/CPE surface was characterized with respect to the effect of pulse parameters. The effect of SW pulse parameters as (frequency f (10 to 140 Hz), scan increment ΔEs (2.0 to 18 mV), and pulse-amplitude a (5.0 to 30 mV)) on the peak current magnitude of 100 nmol L−1 VILD were studied in a BR universal buffer of pH 7.0 following its preconcentration by adsorptive accumulation onto the 5% Ca-MMT/CPE at Eacc = 0.1 V for 40 s. A better-developed peak current magnitude was achieved at f = 80 Hz, ΔEs = 12 mV, and a = 25 mV.

Effect of the accumulation potential and time

As shown in Fig. 11A, the effect of an accumulation potential (Eacc) on the peak current response was investigated from − 0.20 to + 0.5 V after an accumulation time of 40 s for 100 nmol L−1 VILD onto the 5% Ca-MMT/CPE surface. The results were obtained at the optimal conditions at f = 80 Hz, ∆Es = 12 mV, and a = 25 mV in the BR universal buffer of pH 7.0. The peak height increases when the accumulation potential increases from − 0.20 to + 0.1 V, but decreases abruptly up to + 0.5 V onto a 5% Ca-MMT/CPE. The greatest peak current was found at + 0.10 V, according to the plot of stripping peak current as a function of Eacc. Consequently, a potential of + 0.10 V was adopted as the optimum Eacc for VILD detection.

Fig. 11
figure 11

Plots of A SW-AdAS voltammetry peak current (Ip) vs Eacc of 100 nmol L−1 VILD for 40 s in the universal buffer of pH 7.0 and B SW-AdAS voltammetry peak current (Ip) vs. (tacc) of (a) 100, (b) 60, and (c) 9.0 nmol L−1 VILD onto the 5% Ca-MMT/CPE at Eacc = 0.1 V in the universal buffer of pH 7; f = 80 Hz, ΔEs = 12 mV, and a = 25 mV

As shown in Fig. 11B, the influence of accumulation time (tacc) on the response of the electrode in a solution containing three concentration levels of (a) 100, (b) 60, and (c) 9.0 nmol L−1 VILD was investigated. Variation of the tacc showed that the peak current of VILD increased with an increase in the tacc up to 70, 80, and 110 s of (a) 100, (b) 60, and (c) 9.0 nmol L−1 VILD, respectively, then a plateau was observed (Fig. 11B). The peak current was found to increase with increasing the tacc between 0 and 110 s. Above time of 70 s for 100 nmol L−1 VILD, the peak current abruptly decreases and it is almost constant after 80 and 110 s for 60 and 9.0 nmol L−1 VILD, respectively, indicating that the adsorptive saturation of the VILD onto 5% Ca-MMT/CPE surface was achieved. This phenomenon leads to an effective alternative for the increase of the method sensibility. This showed that the amount of adsorbed VILD on the 5% Ca-MMT/CPE surface reached a maximum value at 70 s; hence, a tacc of 60 s was used to avoid electrode saturation. As demonstrated in (Fig. 11B), a tacc of 60 s was used throughout because it combined good sensitivity with a very quick analysis time.

Validation of the analytical method

Linearity

Figure 12A exhibits the SW-AdASV responses of the 5% Ca-MMT/CPE to VILD in the concentrations ranging from 1.0–110 nmol L−1. Figure 12B shows the oxidation peak current enhanced significantly with an increase in the concentration of VILD based on an average of six replicate measurements (n = 6) under optimal conditions.

Fig. 12
figure 12

A SW-AdAS voltammograms of bulk VILD solution of different concentrations equal to (a) 0.0 nmol L−1 (Background, dashed line), (b) 1.0, (c) 3.0, (d) 7.0, (e) 20, (f) 30, (g) 40, (h) 60, (i) 70, (j) 75, (k) 80, (l) 85, (m) 90, (n) 95, (o) 100, and (p) 110 nmol L−1 onto the 5% Ca-MMT/CPE surface in a BR universal buffer of pH 7.0 after an accumulation time of 60 s at the optimal conditions of pulse parameters; f = 80 Hz, ∆Es = 12 mV, and a = 25 mV and B its calibration plot of Ip (µA) vs VILD concentrations (nmol L−1)

The linear relationship of the peak current vs concentrations of VILD at the 5% Ca-MMT/CPE surface was investigated as follow:

$$I_{{\text{p}}} /\mu A\, = \,0.083\, \pm \,1.22\, \times \,10^{{ - 3}} \,C\left( {{\text{nmol L}}^{{ - 1}} } \right)\, + \,0.36\, \pm \,7.91 \times 10^{{ - 3}} ;\,R^{2} \, = \,0.993$$
(9)

The calculated mean detection and quantification limits of VILD in bulk form were 0.285 and 0.950 nmol L−1, respectively (Table S1).

Robustness

The robustness was examined by evaluating the influence of small variations of some of the most important procedure variables, including pH, frequency, preconcentration potential, and preconcentration time (Table S2). The obtained results indicated the reliability of the proposed procedure for the assay of 80 nmol L−1 of bulk VILD; hence, it can be considered as robust. The obtained mean percentage recoveries (R %) and relative standard deviations (RSD %) based on an average of five replicate measurements (n = 5) were not significantly affected within the studied range of variations of some operational parameters. Consequently, the proposed procedure can be considered as robust [69, 70].

Selectivity

The selectivity of the optimized procedures was examined in the presence of some common excipients to monitor the interference effect. The selectivity of the optimized procedure for the determination of VILD was examined in the presence of excipients included in the formulation such as anhydrous lactose and microcrystalline cellulose. Samples containing different concentrations of bulk VILD (without excipients) and different concentrations of the excipient under evaluation were analyzed using the proposed method. The voltammograms of all the tested solutions showed that the peak current and peak potential of VILD did not change (Fig. S2). The obtained mean percentage recoveries R % and RSD % values based on an average of five replicate measurements (n = 5) were 98.90 ± 1.57 to 101.40 ± 0.90 (VILD drug without excipients) and 99.09 ± 1.59 to 101.50 ± 0.87 (with excipients), showed no significant interference from excipients (Table S3). Thus, the proposed procedure can be considered to be specific and selective [69, 70].

Analytical applications

Analysis of VILD in galvus® tablets

The adequacy of the developed method was evaluated by quantifying VILD in its dosage form namely in Galvus® tablets. Precipitation and extraction processes were not required prior to the SWV assay. The corresponding regression equations of previously developed calibration plots were used to calculate the nominal content of the tablet quantity. The accuracy of the method was determined by its recovery using spiked experiments using the calibration curve. The mean results for the determination of VILD using the proposed method were found very close to the declared value of 50 mg. The obtained results showed that the proposed method could be applied with great success to VILD assay in tablets without any interference. Non-significant difference between the slopes of the calibration curves for the VILD drug in bulk form and its commercial formulation at least six times (n = 6) under the same optimal conditions of the linearity.

The mean recoveries in tablets were found to be (99.91–100.04%), with RSD values of (0.96–1.10%). Consequently, the proposed procedure was sensitive and selective for the assay of the VILD in its commercial formulation (Table 1).

Table 1 Assay of VILD in its pharmaceutical formulation (Galvus® tablets) containing 50 mg VILD tablets at 5% (w/w) Ca-MMT/CPE in comparison with the reported gas chromatography–mass spectrometry [10] at (n = 6)

As shown in Table 1 and for showing the accuracy of the proposed method, the results of VILD tablets were compared with the previously reported gas chromatography-mass spectrometry [10].

SWV method was compared with the already published spectrometric method for VILD drug [10], using both Student’s F- and t-tests. At 95% of the confidence level, as shown in Table 1, the F- and t-values were calculated to be 2.41 and 0.37, respectively. These values did not exceed the theoretical F- and t- values at 5.05 and 2.23, respectively. These results show that there were no statistically significant differences between the performances of both methods and also, show the reliability of the previously suggested SW-AdASV method.

Analysis of VILD in spiked human serum

Linearity

To evaluate the applicability of the method to biological samples, VILD was determined in a human serum sample under the same conditions as employed for the pure VILD in bulk form and commercial formulations. As shown in Fig. 13A, anodic stripping voltammograms are illustrated to assay the investigated VILD drug spiked in human serum sample by the method of the calibration curve.

Fig. 13
figure 13

A SW-AdAS voltammograms for the determination of VILD in a spiked human serum at different concentrations equal to (a) 0.0 nmol L−1 (Serum sample, dashed line), (b) 3.0, (c) 5.0, (d) 9.0, (e) 30, (f) 50, (g) 70, (h) 90, (i) 110, (j) 130, and (k) 140 nmol L−1 onto the 5% Ca-MMT/CPE surface in a BR universal buffer of pH 7.0 after an accumulation time of 60 s at the optimal conditions of pulse parameters; f = 80 Hz, ∆Es = 12 mV, and a = 25 mV and B its calibration plot of Ip (µA) vs VILD concentrations (nmol L−1)

As shown in Fig. 13B, the calibration plot of the peak current vs the concentration was observed to be linear over the range 3.0–140 nmol L−1 VILD after an accumulation time of 60 s in a BR universal buffer of pH 7, following the regression equation:

$$I_{{\text{p}}} /\mu A{ }\, = { }\,0.053\, \pm \,{ }2.93{ }\, \times { }\,10^{{{ - }{2}}} { }C{ }\left( {{\text{nmol }}^{{{\text{L}}{ - }{1}}} } \right){ }\, + \,{ }0.392\, \pm \,0.011);\,{ }R^{{2}} { }\, = \,{ }0.999$$
(10)

The VILD peak obtained in serum using 5% (w/w) Ca-MMT/CPE was well shaped and clearly distinguished from the background. The background of the serum showed that no peaks overlapped with the peak of VILD in the studied potential range (Fig. 13A, curve a). For the preconcentration time of 60 s, the obtained mean limits of detection and quantification were of about 0.671 and 2.237 nmol L−1, respectively (Table S1) at least six times (n = 6).

In addition, the applicability of the SW-AdASV procedure for the analysis of VILD in biological samples was also evaluated by estimating its recovery from the spiked human serum sample (Table 2).

Table 2 Assay of VILD in a spiked human serum after an accumulation time of 60 s at 5% (w/w)Ca-MMT/CPE in comparison with the reported chromatographic method [3] (UPLC–MS/MS) at (n = 6)

Within the linearity range, the mean percentage recoveries for the varied concentrations of VILD ranged from (99.90 ± 1.9%) to (101.66 ± 0.92%). Consequently, the suggested approach was suitable for the analytical determination of the VILD drug in a serum sample. The recovery studies of the calibration method on biological samples were carried out to provide further evidence of the validity of the methods. The mean recoveries are in good agreement with the RSD values that are less than 2% (Table 2).

Thus, the precision for the analysis of biological samples is very satisfactory. These results indicate that the content of VILD in the biological samples can be safely determined by using this method without interference from other substances in the blood samples.

Based on the statistical evaluation (F- and t-tests) for these results, there is no significant difference between the results obtained by the developed SW-AdASV procedure and that obtained by the reference method [3]. When comparing the variances of the developed SW-AdASV procedure and the chromatographic reference method [3] (UPLC–MS/MS) at 19.77 nmol L−1 VILD, the F- and t- values were 3.40 and 1.30, respectively. Whereas the calculated F- and t- values were less than the critical values of 5.05 and 2.23, respectively, at the 95% confidence level, confirming the good accuracy of the methods (Table 2).

Effect of foreign species

As shown in Table 3, the influence of various interfering species and co-formulated drugs on the determination of 50 nmol L−1 VILD was studied in biological fluids. The tolerated foreign substances of 50 nmol L−1 VILD were cations, anions, and some other typical co-formulated drugs. A foreign species were considered to cause no interference when it caused a variation of less than ± 5% of the VILD.

Table 3 Interferences from foreign species on analysis of 50 nmol L−1 VILD by the optimized SW-AdASV method utilizing 5% (w/w) Ca-MMT/CPE

The tolerance limits of various species are shown in Table 3. The results indicated that the common coexisting species did not have a significant effect on the determination of VILD. Thus, the method is highly selective and, therefore, has been successfully applied to the trace determination of VILD in various biological samples and human serum without any prior separation or preconcentration steps [37, 71].

Comparison of the proposed method with the reported method

On the other hand, as shown in Table 4, the performance of the 5% Ca-MMT/CPE using SWV for the determination of VILD was compared to that of other techniques and electrodes described in the literature [22, 23]. When analytical data obtained in this study are evaluated, it is seen that the performance of 5% Ca-MMT/CPE is much more sensitive in terms of LOD than the values obtained with PGE and GC/Pt [22, 23]. Based on the determination of VILD in the literature [22], the linear ranges were 2.9–49.9, 2.9–26.4, and (2.9–44.1) × 10−6 mol L−1 in bulk forms, pharmaceutical formulation, and human urine, respectively, and the LOD was 82 nmol L−1 in bulk forms using SWASV. Based on the determination of VILD in the literature [23], the linear range was between (2.0–10) × 10−3 mol L−1 and the LOD was 0.24 × 106 nmol L−1 at a platinum electrode using LSV. As can be seen, the LOD obtained with the proposed electrode (0.28 nmol L−1) is the lowest, and the calibration curve (1.0–110) × 10−9 mol L−1) is the most extensive (Table 4).

Table 4 Comparison between some characteristics of the proposed analytical method for determining VILD with other analytical methods described in the literature

Conclusion

The presented SW-AdASV method was carried out for the determination of VILD using the BR universal buffer at the optimum pH value of 7.0. A new electrochemical method with very high sensitivity has been developed for the detection of VILD on the 5% Ca-MMT/CPE surface. Overall, this modified electrode was sensitive and selective and had a rapid response, and also, the surface renewal was easy. According to our findings, the 5% Ca-MMT/CPE had a high electrocatalytic activity for the oxidation of VILD. In this work, the developed method was applied to determine VILD in bulk form and dosage form, and spiked human serum has the usefulness of being rapid, sensitive, precise, and inexpensive compared to other analytical reported methods. The calculated mean detection limit was 0.671 nmol L−1 in spiked human serum by SWV using 5% Ca-MMT/CPE. The parameters including 0.1 V (Eacc), 60 s (tacc), 80 Hz (f), 12 mV (∆Es), and 25 mV (a) were used and selected as optimum conditions to obtain a well-defined anodic voltammogram for VILD determination. In addition, validation parameters for analytical performance, including linearity, robustness, and selectivity were also evaluated for VILD. Besides, the interference from excipients of VILD does not interfere with the determination. Therefore, extraction procedures are not needed.